DOI:10.2214/AJR.05.0320
AJR 2007; 188:W49-W56
© American Roentgen Ray Society
Hepatic MRI Using the Double-Echo Chemical Shift Phase-Selective Gradient-Echo Technique
Jeong-Sik Yu1,
Jun-Gyun Park1,
Eun-Kee Jeong1,2,
Mi-Suk Park1 and
Ki Whang Kim1
1 Department of Radiology and Research Institute of Radiological Science, Yonsei
University College of Medicine, YongDong Severance Hospital, 146-92
Dogok-Dong, Gangnam-Gu, Seoul 135-720, South Korea.
2 Present address: Department of Radiology, Utah Center for Advanced Imaging
Research, University of Utah, Salt Lake City, UT 84108.
Received February 25, 2005;
accepted after revision October 25, 2005.
Address correspondence to J.-S. Yu
(yjsrad97{at}yumc.yonsei.ac.kr).
WEB
This is a Web exclusive article.
Abstract
OBJECTIVE. The objective of our study was to describe the merits and
drawbacks of the double-echo chemical shift phase-selective gradient-echo
technique for hepatic MRI.
CONCLUSION. With complementary information from two different
dynamic imaging sets in conjunction with errorless subtraction between in- and
out-of-phase images, the double-echo chemical shift phase-selective
gradient-echo technique provides useful information regarding unpredictable
variations in intra- or extralesional lipid content, allowing detailed
assessment of focal lesions during hepatic MRI.
Keywords: abdominal imaging • hepatobiliary imaging • liver disease • MRI • MR technique
Introduction
Adouble-echo chemical shift phase-selective gradient-echo technique allows
simultaneous acquisition of in-phase and out-of-phase images in the multislice
mode on the basis of a spoiled gradient-echo sequence hepatic MRI
[1]. With its intrinsic short
acquisition time and identical slice level for the two different
phase-selective images, this technique affords multiphasic dynamic imaging
[2] and subtraction of
out-of-phase images from in-phase images without registration error
[3]. In this pictorial essay,
we discuss the usefulness of this sequence for standard hepatic MRI and
provide examples that illustrate the advantages and potential drawbacks.
MRI Technique
Chemical shift MRI pulse sequences are used to suppress the signal from
fatty tissue. In contrast to the chemically selective prepulse technique that
saturates the CH2 protons themselves, the phase-selective technique
incompletely refocuses the spin echo, causing the phases of magnetization from
water and CH2 protons to be opposite one another. This results in
destructive interference, and thus a greater sensitivity to a small fraction
of lipid content [4]. The
signal intensity loss is most profound for voxels with similar signal
magnitudes for both water and CH2 and is less for voxels that are
predominantly water or predominantly fat. At the boundary between
predominantly watery and predominantly fatty tissue, however, the individual
voxels contain a substantial water and fat component, and the resultant signal
loss makes black boundary artifact. Different from the chemical shift artifact
depicted on conventional in-phase imaging sequences, the black boundary
artifact is not dependent on the frequency-encoding direction, field of view,
or bandwidth of the pulse sequences.
The calculated ideal TEs for out-of-phase and in-phase images are
approximately 2.2 and 4.5 msec, respectively, when using a 1.5-T unit;
however, because of limited hardware performance, 2.7 and 5.3 msec are
considered to be optimized TEs for out-of-phase and inphase imaging,
respectively, with our machine (Vision, Siemens Medical Solutions) for
double-echo chemical shift phase-selective gradient-echo imaging
[1]. Images are obtained before
and after the injection of an IV contrast agent (gadopentetate dimeglumine
[Magnevist, Schering]) for unenhanced and contrast-enhanced multiphasic
dynamic imaging. Fifteen axial slices are obtained with an 8- to 10-mm slice
thickness, a 1.6- to 2-mm intersection gap, and a 19- to 21-sec acquisition
time, encompassing the entire liver during a single breath-hold. Other
parameters are as follows: TR, 140 msec; flip angle, 90°; and matrix, 128
x 256 at a 6/8 rectangular field of view.
Unenhanced Imaging and In-Phase-Out-of-Phase Subtraction
Diffuse or Focal Hepatic Steatosis
The areas of the liver containing fat show substantial signal intensity
loss on out-of-phase images due to destructive interference between water
protons and CH2 protons
[4]. In-phase-out-of-phase
subtraction can be used to find and delineate a relatively small amount of
lipid in the liver to confirm previously held suspicions (Figs.
1A,
1B, and
1C).
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —43-year-old man with diffuse and geographic hepatic
steatosis. Signal of hepatic parenchyma is homogeneous on transverse in-phase
spoiled gradient-echo MR image (TR/TE, 140/5.3; flip angle, 90°).
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —43-year-old man with diffuse and geographic hepatic
steatosis. Subtracted image of out of phase (B) from in phase
(A) shows high signal intensities of fatty deposition that correspond
to the findings in A and B and that can be distinguished from
dark signal fat-spared areas (arrowheads).
|
|
Focal Fat-Sparing in Fatty Liver
Regardless of the presence of focal hepatic lesions, locally decreased
splanchnic portal venous flow causes diminished fatty infiltration in the
affected area. Fat-spared areas retain their signal intensity and can
therefore be distinguished from the surrounding area of fatty liver on
out-of-phase images, in which signal loss occurs
[5]. Fat-spared areas show no
remaining signal after in-phase-out-of-phase subtraction (Figs.
2A,
2B, and
2C).
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —36-year-old woman with diffuse hepatic steatosis and
metastases from pancreatic cancer. Transverse in-phase spoiled gradient-echo
MR image (TR/TE, 140/5.3; flip angle, 90°) shows multifocal hypointense
lesions (arrows).
|
|
View larger version (118K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —36-year-old woman with diffuse hepatic steatosis and
metastases from pancreatic cancer. Out-of-phase image (TE, 2.7 msec) shows
perilesional hyperintense rims (arrowheads) distinguished from
decreased signal of background fatty liver.
|
|
View larger version (138K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —36-year-old woman with diffuse hepatic steatosis and
metastases from pancreatic cancer. Areas (arrowheads) of dark signal
on subtracted image of out of phase (B) from in phase (A)
imaging includes lesions and perilesional fat-spared areas. Decreased portal
venous perfusion around metastases prevented perilesional fat deposition.
|
|
Focal Hepatic Lesions Containing Lipids
Many hepatocellular lesions contain variable amounts of intracellular
lipids, regardless of their malignant potential
[6]. Despite the high incidence
of lipid content in hepatocellular carcinoma and hepatocellular adenoma, the
presence of small amounts of lipids is not, in itself, indicative of
neoplastic lesions in the liver
[6]. Even a minimal loss of
intralesional signal on out-of-phase images appears hyperintense on
in-phase-out-of-phase subtraction images (Figs.
3A,
3B,
4A,
4B, and
4C). The presence of
intralesional iodized oil after chemoembolization of hepatocellular carcinoma
can also cause signal loss on out-of-phase images because the lipid content is
intermingled with water in the necrotic lesion, thus also showing
hyperintensity after subtraction (Figs.
5A,
5B, and
5C).
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —50-year-old man with cirrhosis from chronic hepatitis B virus
and dysplastic nodule containing intracellular fat. Transverse out-of-phase
image (TR/TE, 140/2.7) shows small hypointense nodule (arrow) in left
lobe of liver. Corresponding in-phase image (not shown) (TE, 5.3 msec) showed
relatively high-signal-intensity nodule at same site.
|
|
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —50-year-old man with cirrhosis from chronic hepatitis B virus
and dysplastic nodule containing intracellular fat. Nodular hyperintensity
(arrow) on subtracted image of out of phase from in phase suggests
presence of fatty component within lesion.
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —55-year-old woman with cirrhosis from chronic hepatitis B
virus and small hepatocellular carcinoma containing small amount of fat.
Transverse in-phase spoiled gradient-echo MR image (TR/TE, 140/5.3 msec; flip
angle, 90°) shows hypointense nodule (arrow). Out-of-phase image
(not shown) (TE, 2.7 msec) also showed hypointensity for same nodular
lesion.
|
|
View larger version (117K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —55-year-old woman with cirrhosis from chronic hepatitis B
virus and small hepatocellular carcinoma containing small amount of fat.
Hyperintense area (arrow) on subtracted image of out of phase from in
phase suggests intralesional fatty content, which was difficult to identify by
direct comparison of in-phase and out-of-phase images.
|
|
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —55-year-old woman with cirrhosis from chronic hepatitis B
virus and small hepatocellular carcinoma containing small amount of fat.
Transverse in-phase dynamic MR image during arterial phase shows lesional
enhancement (arrow) of hypervascular tumor.
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —50-year-old man with cirrhosis and hepatocellular carcinoma
treated by transarterial chemoembolization 3 months earlier. Transverse
in-phase spoiled gradient-echo MR image (TR/TE, 140/5.3; flip angle, 90°)
shows hyperintense nodule (arrow). Profound signal loss was observed
on out-of-phase images (not shown), and subtracted images (not shown) showed
hyperintensity for same lesion due to lipid content of intralesional iodized
oil mixed with necrotic tissue after chemoembolization.
|
|
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —50-year-old man with cirrhosis and hepatocellular carcinoma
treated by transarterial chemoembolization 3 months earlier. During arterial
dominant phase of dynamic MRI, signal intensity of nodule (arrow,
B) is still higher than that of surrounding liver on in-phase image,
mimicking hypervascular lesion; however, opposed-phase image shows low signal
intensity corresponding to nonenhancement (arrow, C) of
necrotic tumor.
|
|
View larger version (120K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C —50-year-old man with cirrhosis and hepatocellular carcinoma
treated by transarterial chemoembolization 3 months earlier. During arterial
dominant phase of dynamic MRI, signal intensity of nodule (arrow,
B) is still higher than that of surrounding liver on in-phase image,
mimicking hypervascular lesion; however, opposed-phase image shows low signal
intensity corresponding to nonenhancement (arrow, C) of
necrotic tumor.
|
|
Drawbacks of Out-of-Phase Imaging and In-Phase-Out-of-Phase Subtraction Imaging
For lesions containing excessive amounts of lipid, including rare
lipomatous lesions or severe fatty metamorphosis in hepatocellular carcinomas,
no intralesional signal loss can be observed on the corresponding out-of-phase
images similar to subcutaneous or mesenteric fat. Intravoxel cancellation
artifacts at the interface between the excessive fat and water content on
out-of-phase images create a hyperintense rim present on in-phase-out-of-phase
subtraction images. This hyperintense rim is useful for distinguishing gross
intralesional fat from background parenchyma with excessive water content
(Figs. 6A,
6B,
6C,
6D, and
6E). This technique might also
be used to improve the evaluation of peripheral tumors for possible
extravisceral extension or for the determination of visceral contours. For
small subcapsular lesions, however, the intravoxel phase cancellation at the
interface between the liver and peritoneal fat can obscure the lesion itself
and lead to erroneous conclusions.
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —43-year-old woman with benign lipomatous tumor in fatty
liver. Transverse in-phase spoiled gradient-echo MR image (TR/TE, 140/5.3;
flip angle, 90°) shows lobulated hyperintense lesion (arrow) in
right lobe of liver.
|
|
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —43-year-old woman with benign lipomatous tumor in fatty
liver. Out-of-phase image (TE, 2.7 msec) shows dark rind of intravoxel phase
cancellation at periphery of lesion and at interface between fatty mass and
surrounding hepatic parenchyma (arrow). Centrally preserved
hyperintensity originates from excessive proportion of fatty content, which is
comparable to subcutaneous fat. Geographically decreased signal on background
of hepatic parenchyma is due to hepatic steatosis.
|
|
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6D —43-year-old woman with benign lipomatous tumor in fatty
liver. In-phase arterial phase dynamic MR image shows homogeneously high
signal intensity (arrow), suggesting diffuse contrast enhancement of
lesion.
|
|
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6E —43-year-old woman with benign lipomatous tumor in fatty
liver. Homogeneously decreased signal of lesion (arrow) on
out-of-phase arterial phase image suggests paradoxical decrease in signal
intensity for enhancing lesions containing excessive fatty component.
Intralesional attenuation was approximately -50 H on CT scan (not shown), and
nonmalignant cellular fibrosis with abundant fat globules was verified by
percutaneous gun needle biopsy.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7A —38-year-old man with advanced cirrhosis containing
innumerable siderotic nodules. Transverse out-of-phase spoiled gradient-echo
MR image (TR/TE, 140/2.7; flip angle, 90°) shows contracted hepatic
parenchyma with surface nodularity (arrowheads) surrounded by massive
ascites (asterisk) and subcapsular slightly hyperintense lesion
(arrow).
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7B —38-year-old man with advanced cirrhosis containing
innumerable siderotic nodules. In-phase image (TE, 5.3 msec) shows
high-signal-intensity lesion (arrow) well distinguished from
background parenchyma containing innumerable dark-signal-intensity nodules.
Darkened signal of siderotic nodules is due to T2* effect with
longer TE of in-phase imaging.
|
|
View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7C —38-year-old man with advanced cirrhosis containing
innumerable siderotic nodules. Subtracted image of out of phase from in phase
(B - A) shows "blackout" of hepatic parenchymal signal
(arrowheads) due to negative signal value after subtraction. In this
situation, presence of fatty component in lesion or background hepatic
parenchyma cannot be properly determined. Arrow = subcapsular focal lesion,
asterisk = ascites.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8A —62-year-old man with chronic hepatitis B virus and diffuse
hepatic steatosis complicated by small hepatocellular carcinoma. Transverse
in-phase spoiled gradient-echo MR image (TR/TE, 140/5.3; flip angle, 90°)
shows slightly hypointense nodule (arrow).
|
|
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8B —62-year-old man with chronic hepatitis B virus and diffuse
hepatic steatosis complicated by small hepatocellular carcinoma. Lesion is not
well delineated on out-of-phase image (TE, 2.7 msec) because of signal loss
from background liver as result of diffuse fat infiltration and consequently
decreased lesion-to-liver contrast.
|
|
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8C —62-year-old man with chronic hepatitis B virus and diffuse
hepatic steatosis complicated by small hepatocellular carcinoma. Transverse
in-phase (C) and out-of-phase (D) dynamic MR images during
arterial phase show iso- and mild hyperintensity of lesion, respectively
(arrows); these findings suggest hypervascular tumor when compared
with unenhanced images. Lesion-to-liver contrast is greater on D than
C due to signal loss from background hepatic parenchyma on
D.
|
|
View larger version (137K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8D —62-year-old man with chronic hepatitis B virus and diffuse
hepatic steatosis complicated by small hepatocellular carcinoma. Transverse
in-phase (C) and out-of-phase (D) dynamic MR images during
arterial phase show iso- and mild hyperintensity of lesion, respectively
(arrows); these findings suggest hypervascular tumor when compared
with unenhanced images. Lesion-to-liver contrast is greater on D than
C due to signal loss from background hepatic parenchyma on
D.
|
|
For cases of hepatic iron or copper deposition, the TE of in-phase imaging
is long enough for T2* relaxation. The resultant signal loss on
in-phase images can lead to inappropriate in-phase-out-of-phase subtraction,
which results in a negative value for the remaining signal. In this situation,
there is a chance for false-negative or false-positive findings. Examples
include false-negative findings such as the lipid component intermingled with
iron or copper deposition or false-positive findings such as non-fatty focal
lesions in the background of signal blackout (Figs.
7A,
7B, and
7C).
IV Contrast-Enhanced Dynamic Imaging
Focal Lesions Within Nonfatty Liver
Regardless of the intralesional lipid fraction, using unenhanced images as
a reference, a signal increase observed during the arterial phase dynamic
imaging is suggestive of a hypervascular tumor (Figs.
4A,
4B, and
4C). Identifying
hypervascularity on in-phase images is sometimes difficult because of the
inherently high signal intensity of fat-containing lesions on unenhanced
images that could show signal loss on the corresponding out-of-phase images
(Figs. 5A,
5B, and
5C).
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8E —62-year-old man with chronic hepatitis B virus and diffuse
hepatic steatosis complicated by small hepatocellular carcinoma. Transverse
in-phase (E) and out-of-phase (F) 5-minute delayed dynamic MR
images show decreased signal due to washout of contrast agent
(arrows) from lesion. Fibrotic pseudocapsule around lesion is better
delineated on F due to signal loss of background parenchyma.
|
|
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8F —62-year-old man with chronic hepatitis B virus and diffuse
hepatic steatosis complicated by small hepatocellular carcinoma. Transverse
in-phase (E) and out-of-phase (F) 5-minute delayed dynamic MR
images show decreased signal due to washout of contrast agent
(arrows) from lesion. Fibrotic pseudocapsule around lesion is better
delineated on F due to signal loss of background parenchyma.
|
|
Theoretically, there is a risk of paradoxical signal suppression on
out-of-phase images after contrast injection for lesions with excessive lipid
content [7] (Figs.
6A,
6B,
6C,
6D, and
6E). In daily practice,
however, the amount of lipid in fatty liver or hepatocellular nodules is
relatively small, which minimizes this problem. Exceptions include extreme
cases of fatty liver; fatty tumors including lipoma, liposarcoma, and
angiomyolipoma; or gross fatty metamorphosis of hepatocellular carcinomas with
negative attenuation (in Hounsfield units). To avoid misinterpretation,
combined observations involving both out-of-phase and in-phase dynamic imaging
are essential to characterize the temporal enhancement pattern of
fat-containing lesions.
Focal Lesions in the Background of Hepatic Steatosis
For hypointense lesions on unenhanced T1-weighted images, the visual
conspicuity of lesions in the fatty liver is poor on out-of-phase images. This
is due to the signal loss from background parenchyma, resulting in decreased
lesion-to-liver contrast (Figs.
8A,
8B,
8C,
8D,
8E, and
8F). Although background
signal suppression is generally beneficial for the detection of hypervascular
foci within the fatty liver, the enhancing foci, which are not detected on
unenhanced images, have the potential to be misinterpreted as pseudolesions
from nontumorous focal perfusion variations
[8]. Because of unpredictable
variations in lesion-to-liver contrast on the out-of-phase images, in-phase
imaging can be more accurate for determining the vascular nature of focal
lesions in a background of hepatic steatosis.
Conclusion
In unenhanced imaging, the double-echo chemical shift phase-selective
gradientecho technique is more helpful for the detection of small amounts of
lipids than visual inspection of in-phase and out-of-phase images alone
because it benefits from the automatic in-phase-out-of-phase subtraction
without slice misregistration. Combined interpretation of out-of-phase and
inphase IV contrast-enhanced dynamic images provides useful information for
determining the temporal enhancement characteristics of focal lesions without
incurring the problem of unpredictable variations in intra- or extralesional
lipid content.
References
- Taupitz M, Deimling M, Malcher R, Schoroter T, Bollow M, Hamm B. A
new rapid T1-weighted multiplanar spoiled gradient-echo sequence for
simultaneous acquisition of in-phase and opposed-phase images (SINOP). (abstr)
In: Proceedings of the International Society of Magnetic Resonance
in Medicine. Berkley, CA: ISMRM, 1998:517
- Noguchi Y, Murakami T, Kim T, et al. Detection of hypervascular
hepatocellular carcinoma by dynamic magnetic resonance imaging with
double-echo chemical shift in-phase and opposed-phase gradient echo technique:
comparison with dynamic helical computed tomography imaging with double
arterial phase. J Comput Assist Tomogr2002; 26:981
-987[CrossRef][Medline]
- Honjo K, Murata K, Takizawa O, et al. Optimization of SINOP
sequence with automatic subtraction (IP-OP) for liver imaging. (abstr) In:
Proceedings of the International Society of Magnetic Resonance in
Medicine. Berkley, CA: ISMRM, 1999:517
- Mitchell DG, Kim I, Chang TS, et al. Fatty liver: chemical shift
phase-difference and suppression magnetic resonance imaging techniques in
animals, phantoms, and humans. Invest Radiol1991; 26:1041
-1052[Medline]
- Chung JJ, Kim MJ, Kim JH, Lee JT, Yoo HS. Fat sparing of
surrounding liver from metastasis in patients with fatty liver: MR imaging
with histopathologic correlation. AJR2003; 180:1347
-1350[Abstract/Free Full Text]
- Nakanuma Y, Hirata K, Terasaki S, Ueda K, Matsui O. Analytical
histopathological diagnosis of small hepatocellular nodules in chronic liver
diseases. Histol Histopathol1998; 13:1077
-1087[Medline]
- Mitchell DG, Stolpen AH, Siegelman ES, Bloinger L, Outwater EK.
Fatty tissue on opposed-phase MR images: paradoxical suppression of signal
intensity by paramagnetic contrast agents. Radiology1996; 198:351
-357[Abstract/Free Full Text]
- Yu JS, Kim KW, Jeong MG, Lee JT, Yoo HS. Nontumorous hepatic
arterial-portal venous shunts: MR imaging findings.
Radiology2000; 217:750
-756[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
